Audio Question Answering on Clotho (test)
52.8Token-Level AccuracyGST-T2
Evaluation results
| Method | Links | |
|---|---|---|
| GST-T2Training Strategy=Full-Data Training, T.E.=~2d2026.05 | 52.8 | |
| SALMONNTraining Strategy=Full-Data Training, T.E.=N/A2026.05 | 52.4 | |
| GST-G3Training Strategy=Full-Data Training, T.E.=~2d2026.05 | 52.4 | |
| GST-G2Training Strategy=Full-Data Training, T.E.=~2d2026.05 | 51.9 | |
| Mix AllTraining Strategy=Full-Data Training, T.E.=~4d2026.05 | 50.6 | |
| IndividualTraining Strategy=Full-Data Training, T.E.=~7d2026.05 | 47.5 | |
| SequentialTraining Strategy=Full-Data Training, T.E.=~7d2026.05 | 47 | |
| GST-T2Training Strategy=Low-Resource Finetuning, T.E.=~0.5d2026.05 | 46.9 | |
| GST-G3Training Strategy=Low-Resource Finetuning, T.E.=~1d2026.05 | 46.5 | |
| Mix AllTraining Strategy=Low-Resource Finetuning, T.E.=~0.5d2026.05 | 46.3 | |
| GST-G2Training Strategy=Low-Resource Finetuning, T.E.=~0.5d2026.05 | 46.1 |